in trying to solve another problem, I’ve been making a lot of sample infusion parts so I can see what’s going on and maybe solve my aluminum mold part issues. I’ve been doing infusions on lexan with inf-114-210 epoxy, and I always always get the same problem, which is a surface where the area between the weave is not filled with resin, I’ve tried lowering vacuum pressure, different mold release, peel ply, no peel ply, etc. now I’m wondering if it might be possible that the capillary action of the fibers drawing resin into them might be stronger than the vacuum and is sucking resin into them and away from the surface, in this case, should I try throttling the resin flow to as slow as I can infuse it without it curing in the pot?
Are you doing a vacuum check? Whats the vacuum drop rate? Do you degass the resin? Do you have photos of your setup?
Looks dry to me.
Always, always carry out a vacuum integrity test prior to infusion.
No measurable vacuum drop, degassed the resin, noticed bubbles forming when the resin hit the vacuum though, did another test today with holding vacuum on the stack for 15 minutes then infusing with a restrictor at 12 psi instead of 14.7 we’ll see how it turns out
I would recommend slowing your infusion by restricting the resin flow. I always infuse with max vacuum. If the resin is flowing too quickly over the top of the stack, it can cause similar problems. I always use 1/4" OD supply lines and even then sometimes have to restrict flow to ensure the full stack get saturated. One more thing that may have nothing to do with your problem is to keep your supply source 2 to 3 feet below the mold.
Agree on the speed issue. Slow the infusion rate of the epoxy would be my first guess. Get some line clamps or something to squeeze off the line.
I also agree on trying to slow your infusion speed. The more layers in your laminate stack, the longer it will take the resin to penetrate to the mold surface. Ideally, the resin front should be wetting the laminate all the way to the mold surface at about the same speed it travels across the surface. The resin will fill in a bit if the resin on the surface hits the vacuum line first, but only to a point.
Peel ply breaks (brakes); however you spell it, can serve to slow infusion toward the end and allow the resin to penetrate fully before hitting the vacuum line. You can also clamp the feed line periodically to allow saturation to catch up to the flow across the surface.
Here’s a video of an infusion I did for one skin of a foam core panel. The laminate is 3 layers of plain weave (9 oz./yd) carbon fiber. Polyester peel ply on the surface, then flow media (which is cut to slightly smaller than the laminate). The resin is West Systems 105/206. There is also a pause in about the middle of the process to mix more resin. The laminate was completely free of voids, dry spots and pinholes.
I finally got a good test panel with 12psi of vacuum and thoroughly degassed resin, it definitely takes longer to saturate carbon than it does fiberglass which is what I am used to working with. I also rethought my aluminum molds and am going to redesign them to allow for an imperfect vacuum seal, I watched the infused falcon supercar manifold video a few times and realized their mold uses the vacuum to clean up any leaks in the perimeter seal, otherwise they wouldn’t be able to get away with o rings.
Why do you ask for a theory? It is physic, not more or less.
For a perfect infusion you need a 100% tight mould and bag, degassed resin and a a perfect vacuum, which means less then 20mbar absolute pressure.
You can read that in the MTI thread.
there’s perfect in physics, not in engineering. I doubt any mold has a “perfect” vacuum, nor is any resin perfectly “degassed”, there appear to be elements in the setup and process that allow for some amount of imperfection in the vacuum seal, trapped gas and volatiles, without actually amounting to a flawed part. when I tested my aluminum mold with butyl seal, there was no vacuum leak detectable, then I removed the laminate stack, and inflated the bladder, basically turning it into an incredibly low volume of vacuum, then, I did see about 1psi per 10 minutes drop in that incredibly small volume (maybe 5cc). so it looks like the mold really needs to be designed in a way that the vacuum channels scavenge the perimeter seal and keep any gas flowing to the vacuum pump before getting to the resin
You really should use a digital absolute pressure gauge (such as the GDH200-14) at the resin inlet to measure the pressure change over a period of 10mins or more. Using analog gauges are not accurate.
did some research and also some experimentation. it seems to me that capillary action in resin infusion has a big effect, also, the effect is probably more pronounced under compressed bladder mold infusion, which might be relevant to some people on here. it would seem that the capillary action in carbon fiber would increase in strength as the fibers are closer together, so the effect I was seeing where the carbon would wick resin and create dry spots even with fully degassed resin and no vacuum leaks, would actually be amplified by further compacting the dry stack. this was observed in infusion tests using aluminum molds and 120psi bladder pressure. the effect was decreased and eliminated by lowering the infusion vacuum level, and decreasing the bladder pressure. would seem to me that prepregs wouldn’t see this problem because the fibers are already fully saturated before going into the mold… will keep experimenting going
1psi per 10 min is not a small leak, thats a much too big leak for infusion! That is 6.8mbar per minute, maximum is 1mbar per min on bigger parts, most we have 0.1mbar/min droprate!
my vacuum scavenge channel cleans up an imperfect perimeter seal, so what I’m observing now is no bubbles or evidence of gas in the laminate, and am nailing down the surface dryness issue, which correlates to how long the laminate is wetted out before the resin gels, the longer, the better the results
It is a vacuum and setup problem.
did an interesting test, instead of using vacuum to pull resin into the bladder molded part, I built a pressure pot and injected it. and found bubbles coming from the overflow port, assumed it was air being pushed out from the incoming resin, but the bubbles didn’t stop when pressure in the pot was released, then they came up from the input line. I think what I learned because there was no vacuum involved to pull air into the mold, is that the bladder, either latex, or epdm rubber, was bleeding air into the laminate, some reading online led to some studies on the gas permeability of epdm and latex, and a lot of other materials, all of which will let gas through proportional to pressure and inverse to thickness… I’m trying another test now with a vacuum bag sleeve and peel ply inside it surrounding the latex bladder, that way I’m assuming any air bleeding through will escape to the side openings rather than into the laminate. so far results look good with no bubbles in the overflow or feed lines